Publikationen

Autor	Zelle, Daniel
Datum	2024
Art	Thesis
Betreuer	Waidner, Michael
Abstrakt	This dissertation focuses on the cybersecurity of modern electric vehicles (EVs), specifically their communication protocols. The research highlights IT security risks, such as the manipulation of driving behavior, data breaches during the charging process, and threats to the stability of the power grid from EV charging. A threat and risk analysis (TARA) is conducted to identify the critical points of attack in vehicle communication. We present an innovative approach to automating the assessment of the attack surface, which will accelerate the TARA process and reduce errors. The assessment reveals high risks in internal vehicle communication as well as in charging communication. To secure internal communication via the CAN bus, we propose the BusCount protocol as a secure communication solution for protecting automotive networks. This approach provides clear advantages over existing solutions in protecting against replay and delay attacks. Furthermore, we investigate securing modern automotive Ethernet using TLS and validate various communication scenarios. We assess the potential performance impacts of different ciphers on typical automotive hardware and compare them to the industry’s requirements. Security gaps were identified in the widespread SOME/IP protocol in service-oriented communication via automotive Ethernet. We developed two possible protocol extensions to secure SOME/IP. This work also addresses potential manipulation of the power grid and data protection issues during the charging process to ensure the security of the EV charging infrastructure. We show how to reduce the transmission of personal data during the charging process and propose an extension for the existing Plug & Charge protocols. Our proposed solution utilizes a Direct Anonymous Attestation (DAA) scheme for anonymous charging. In summary, this work contributes to advancing EV cybersecurity by identifying critical aspects through a risk assessment and proposing ways to secure internal Controller Area Network (CAN) and automotive Ethernet communication. Additionally, this work addresses security and privacy issues related to the EV charging infrastructure. These findings and solutions provide a solid foundation for creating a more secure environment for EVs in the rapidly evolving automotive industry.
Url	https://publica.fraunhofer.de/handle/publica/479149